econstor www.econstor.eu

Der Open-Access-Publikationsserver der ZBW – Leibniz-Informationszentrum WirtschaftThe Open Access Publication Server of the ZBW – Leibniz Information Centre for Economics

Nutzungsbedingungen:Die ZBW räumt Ihnen als Nutzerin/Nutzer das unentgeltliche,räumlich unbeschränkte und zeitlich auf die Dauer des Schutzrechtsbeschränkte einfache Recht ein, das ausgewählte Werk im Rahmender unter→ http://www.econstor.eu/dspace/Nutzungsbedingungennachzulesenden vollständigen Nutzungsbedingungen zuvervielfältigen, mit denen die Nutzerin/der Nutzer sich durch dieerste Nutzung einverstanden erklärt.

Terms of use:The ZBW grants you, the user, the non-exclusive right to usethe selected work free of charge, territorially unrestricted andwithin the time limit of the term of the property rights accordingto the terms specified at→ http://www.econstor.eu/dspace/NutzungsbedingungenBy the first use of the selected work the user agrees anddeclares to comply with these terms of use.

zbw Leibniz-Informationszentrum WirtschaftLeibniz Information Centre for Economics

Plachta, Robert C.

Working Paper

Fiscal Equalisation and the Soft BudgetConstraint

Finanzwissenschaftliche Diskussionsbeiträge / FinanzwissenschaftlichesForschungsinstitut an der Universität zu Köln, No. 08-8

Provided in cooperation with:Universität zu Köln

Suggested citation: Plachta, Robert C. (2008) : Fiscal Equalisation and the Soft BudgetConstraint, Finanzwissenschaftliche Diskussionsbeiträge / FinanzwissenschaftlichesForschungsinstitut an der Universität zu Köln, No. 08-8, http://hdl.handle.net/10419/27681

FiFo-CPE Discussion Papers Finanzwissenschaftliche Diskussionsbeiträge

Discussion Paper No. 08-8

Fiscal Equalisation and the Soft Budget Constraint

Robert C. Plachta

2008

Finanzwissenschaftliches Forschungsinstitut an der Universität zu Köln

FiFo Köln is a Member of CPE - fifo-koeln.de Cologne Center for Public Economics cpe.uni-koeln.de

http://fifo-koeln.de http://cpe.uni-koeln.de

FiFo-CPE Discussion Papers Finanzwissenschaftliche Diskussionsbeiträge

No. 08-8 / November 2008

ISSN 0945-490X

Fiscal Equalisation and the Soft Budget Constraint

Dipl.-Volksw. Robert C. Plachta1

1 plachta@wiso.uni-koeln.de.

Finanzwissenschaftliches Forschungsinstitut an der Universität zu Köln

A Member of CPE - Cologne Center for Public Economics

P.O. Box 420520, D-50899 KÖLN Zülpicher Str. 182, D-50937 KÖLN

T. +49-221-426979 F. +49-221-422352

Fiscal Equalisation and the Soft Budget Constraint

Robert C. Plachta �

Cologne Center for Public Economics

November 2008

Abstract

This paper assesses the interactions of horizontal �scal equalisation schemes with

debt policy by sovereign regional governments. Local public goods are either �-

nanced by debt or taxation. A horizontal equalisation scheme eleviates regional

public revenue disparities under horizontal and vertical tax competition. We show

that �scal equalisation schemes have no impact on the optimal central government

grant whereas they can either soften or harden the regional budget constraint de-

pending on the speci�c formulae. Revenue equalisation softens the budget constraint

whereas tax base equalisation hardens the budget constraint of poor states.

Keywords: Fiscal federalism, public debt, soft budget constraint,

�scal equalisation, tax competition

JEL Codes: E62, H7

�plachta@wiso.uni-koeln.de. Cologne Center for Public Economics, University of Cologne,Albertus-Magnus-Platz, 50923 Cologne, Germany.

1 Introduction

This paper analyses the impact of horizontal �scal equalisation schemes on both

regional and national budgetary decisions in a federation. A soft budget problem

arises when opportunistic behaviour of subnational governments, who strive for

higher grants from the federal budget, meets the inability of the federal government

to deny payments. According to Rodden et al. (2006), the soft budget problem is

inherent to most federal regimes. Extent and course of action concerning defaulting

jurisdictions depend on the speci�c institutions. In this paper, we turn to a spe-

ci�c institution present in many federal regimes: a �scal equalisation scheme, which

institutionalises horizontal payments for interregional revenue redistribution. We

analyse the interdependency of a horizontal equalisation scheme and the incentive

of decentralised jurisdictions to �nance public goods through debt.1

Financial equalisation is a constituent part of most federations. Beyond that,

their subnational states are endowed with autonomous debt issuing rights. Making

continuous use of debt �nancing goes along with increased interest payments and

may cause budgetary distress. If a supporting jurisdiction comes to aid by granting

bailouts, a soft budget problem emerges. Opportunistic local governments could

be induced to reduce the tax burden of their own resources and to �nance public

goods by debt; con�dent that national government will grant payments.2 A �scal

equalisation scheme deters tax setting by subnational governments. This a¤ects the

incentive to borrow and �nance public goods through de�cits.

This paper consists of four sections and a conclusion. The following section 2 de-

scribes the basic model and the reference scenario. Thereafter all three equalisation

schemes are subsequently analysed in sections 3 to 5.

The concept of the soft budget phenomenon dates back to Kornai (1979).3 While

Goodspeed (2002) already uncovered a �common pool e¤ect� when bailout pay-

ments to subnational governments are �nanced by taxes a¤ecting all jurisdictions,

he neglected tax interactions amongst jurisdictions. Building on Goodspeed�s de-

centralised leadership model Breuillé et al. (2006) merge vertical and horizontal

tax competition in line with Keen and Kotsogiannis (2002) creating a soft-budget-

1The term �institutionalized �scal equalization�marks horizontal payments amongst decen-tralized jurisdictions, which are directly linked to speci�c �nancial resources of the state.

2The term �discretionary� grants marks vertical payments from the national to regional go-vernments, which are not linked to speci�c �nancial resources of the state.

3Oates (2005) o¤ers a short literature review of the development of research on the soft budgetproblem. See Vigneault (2007) for a literature review on the matter within the �scal federalismstudies.

1

framework. They show that tax interactions have no impact on the optimal central

government grant allocation but they harden the budget constraint when regional

tax base outweighs regional debt.

This paper intends to contribute to the literature on soft budgets and �scal insti-

tutions. Building on the framework of Breuillé et al. (2006) we incorporate two di¤er-

ent schemes of horizontal �scal equalisation suggested by Köthenbürger (2002) and

a scheme suggested by Groneck and Plachta (2008). Köthenbürger (2002) analysed

the e¢ ciency aspects of �scal equalisation in a classic horizontal tax competition

framework in line with Zodrow and Mieszkowski (1986) or Wilson (1986). He shows

that tax revenue equalisation exacerbates ine¢ cient tax setting.4 Bucovetsky and

Smart (2002) and (2006) could maintain another result of Köthenbürger in a more

general setting. Tax base equalisation can induce e¢ cient taxation by sovereign

regional governments. The model presented here uncovers the incentive e¤ects of

borrowing induced by a �scal horizontal equalisation scheme with distorting tax

setting. The paper takes account of three di¤erent �scal equalisation schemes:

� Revenue equalisation, representing Germany�s constitutional design;

� Tax base equalisation, representing Canada�s federal-provincial �scal arrange-ments; and

� Expenditure equalisation

2 The base model

The course of action is based on the two period model similar to Goodspeed (2002).

Regions �nance the provision of public goods in the �rst period by �xed grants from

the central government and additionally by debt. In the second period debt has

to be amortized. Regions act as Nash competitors for mobile capital. Each region

simultaneously sets tax rates on capital as in Keen and Kotsogiannis (2002). Decen-

tralized leadership lets the central government act after local governments have made

their decision upon taxes and debt.5 The central government pursues his own goals.

Local governments are involved in a �scal horizontal equalisation scheme, which

speci�c form is decisive for the softness of the regional budget constraint. Three

4Cf. Baretti et al. (2002).5Decentralized leadership captures best the co-determination of national policies through lower

and upper house in federally organized democracies.

2

di¤erent institutional arrangements are analyzed: Revenue equalisation, represent-

ing Germany�s constitutional design; tax base equalisation, representing Canada�s

federal-provincial �scal arrangements; and expenditure equalisation. The central

government can grant payments to the regions to realises their own goals. In con-

trast to Goodspeed (2002) we explicitly account for the central government�s need

to generate revenue to �nance transfers.

The household The representative household in region i = 1; :::; n gains utility

from the consumption of local public goods (gi1; gi2) and private goods (ci1; ci2) in

two periods:

Ui (ci1; ci2; gi1; gi2) = ui (Gi1) + �ui (Gi2) + wi (ci1) + �wi (ci2) : (1)

The subutility functions ui (:) and wi (:) ful�ll the usual characteristics, marginal

utility is positive and diminishing with increasing consumption. Parameter � 2 [0; 1]denotes a discount factor of second period utility. Households are endowed with

exogenous income wi at the beginning of the �rst period, which is spent either for

private consumption ci1 or saved in Si = sii +P

j 6=i sji :

wi = ci1 + Si: (2)

Household savings sji of region i are invested in j, sii denotes savings invested in the

home region. In the second period savings and interest ri are taxed by regional tax

� i and national tax � c. The sum of all savings in region j account to Sj =nPi=1

sji .

Private consumption in the second period amount to

ci2 =

nXj=1

(1 + rj � � j � � c) sji : (3)

Households are interregional immobile. Taxes are levied according to the source

principle.

The capital market The capital market is modelled according to Keen and Kot-

sogiannis (2002). Capital is mobile between regions. Private savings are determined

endogenously depending on the capital supply and demand functions. Capital sup-

ply depends on the production possibilities. The representative �rm in region i

produces output F (Ki) using the sum of savings within the region Ki =P

j sij.

The production function F (:) is strictly concave and twice di¤erentiable. Pro�t

3

maximisation yields the arbitrage condition F 0 (Ki) = ri: The condition implicitly

determines capital demandKi = Ki (ri), which depends negatively on the before-tax

interest rate ri. Due to perfect capital mobility the net return � is equalised across

regions:

� = ri � � i � � c = rj � � j � � c 8i; j: (4)

Private savings Si (�) along with capital supply is determined through utility

maximisation by households:

maxUi = ui (Gi1) + �ui (Gi2) + wi (ci1) + �wi (ci2) (5)

s.t.

wi = ci1 + Si

ci2 = (1 + �)Si:

The consumption-savings decision is made under given taxes and public good bundle.

The optimality condition reads:

@wi@ci1

� @wi@ci2

= (1 + �) : (6)

As usual, the marginal rate of substitution between private consumption in �rst and

second period must match the marginal rate of transformation.

Capital market clearing requires balancing capital demand and supply:

nXi=i

Ki (ri) =nXi=i

Si (�) : (7)

The in�uence of tax policy on net return is of great importance within the model.

Discussing some aspects in advance eases reading, so we �nd that net return depends

negatively on regional and national capital taxation.

d�

d� i=

K0iPn

i=i S0i (�)�

Pni=iK

0i

2 [�1; 0] ; (8)

d�

d� c=

Pni=iK

0iPn

i=i S0i (�)�

Pni=iK

0i

2 [�1; 0] : (9)

A rise in the tax rate does not reduce net return in equal amount. Increasing capital

costs leads to lower demand, hence marginal product and before-tax interest rate

4

rise.drid� i

= 1 +d�

d� i2 [0; 1] ; (10)

drid� c

= 1 +d�

d� c2 [0; 1] : (11)

Horizontal and vertical tax externalities of a common tax base work through the

uniform net return on capital �:

The regional government government activity on the regional level consists of

imposing taxes, issuing debt, and supplying public goods. The regional government

maximises utility of the representative household. In the �rst period the govern-

ment has an exogenous revenue Ti1. Financing further provision of public goods

Gi1 is possible through de�cit Bi1:6 The region can be involved in an horizontal

�scal equalisation scheme �1. Grants (contributions) to region i are denoted by

�i1 > (<)0. The budget constraint in the �rst period is

Gi1 = Ti1 +Bi1 + �i1 (�) : (12)

In the second period a regional government also provides a public good Gi2.7 At

the same time the government must repay debt including interest payments due

(1 + ri)Bi1. Hence, the intertemporal budget must be balanced. To �nance pay-

ments and public goods a region can tax capital (� iKi) : Beyond that the region

(possibly) receives grants Ti2 from the national government. The regional budget

can be involved in a horizontal �scal equalisation scheme �2: Consistently grants

(contributions) to region i are denoted by �i2 > (<)0. The second period regional

public budget is

Gi2 + (1 + ri)Bi1 = Ti2 + � iSi + �i2 (�) : (13)

The national government government activity on the national level consists

solely of issuing grants to regions �nanced by a tax on capital � c in all regions. The

national government is benevolent, maximising the utilitarian welfare function

nXi=1

Ui: (14)

6By simpli�cation, debt is held by foreign investors, cf. Breuillé et al. (2006) and (2007), p. 8.7All public goods in this model are local by nature, that is they exhibit no spillovers.

5

The grants to regions Ti2 8i serve as instruments in the second period.8 The go-vernment budget is

nXi=1

Ti2 = � c

nXi=1

Si = � c

nXi=1

nXj=1

sji : (15)

Horizontal tax competition for mobile capital a¤ects governments to undersupply

public goods as in Zodrow and Mieszkowski (1986). In this model vertical tax

competition arises due to the national government occupying a common tax base.

An increase of the regional or national tax rate varies the tax base of all jurisdictions.

2.1 The sequence of action

The sequence of action of regional and national governments and households is

parted in three stages. On the �rst stage regional governments decide upon the

level of public goods provided in the �rst period. Regions act as Nash competitors.

The only instrument available for �rst period public consumption is debt. Regional

governments are �rst movers to the national government, that is regions perfectly

anticipate national government�s subsequent action. The national government de-

cides on stage two upon payments to regions, henceforth called vertical grants. This

decision requires setting the national tax rate on capital. Simultaneously regions

choose their tax rate for public good provision in the second period. All jurisdic-

tions therefore act as Nash competitors. When setting tax rates, regions can be

involved in a horizontal equalisation scheme speci�ed below. They perfectly an-

ticipate the functioning of the scheme. On the last stage households meet their

consumption-savings decision.

The model is solved using backwards induction. The household�s decision has

been deduced above. Subsequently we turn to stage two to disclose the national go-

vernment�s incentive to grant payments. Concluding we uncover the opportunistic

behaviour of the regional government. In doing so we di¤erentiate cases without

and with speci�c equalisation schemes.

2.2 The incentive to bail out

How tax competition of decentralised jurisdictions can harden the budget constraint

has been uncovered by Breuillé et al. (2006).9 We shortly review their insights, that

8This model is consistent with Breuillé et al. (2006). Some minor mistakes are amended, suchas the imprecise distinction between regional savings Si and regional capital supply Si:

9Cf. Qian and Roland (1998), who discuss the soft budget problem within a simple model ofjurisdictional decentralisation without tax competition.

6

is to begin with, there is no equalisation scheme: �i1 = �i2 = 0 8i. This scenariowill serve as the base case for the following analysis. A soft budget constraint

arises, when the national government responds to increased debt �nanced provision

of public goods through regional governments by increasing vertical grants. The

inability of the national government to commit to a no-bailout-policy is at the heart

of the soft budget phenomenon. A region�s decision upon �scal policy is deterred if it

can expect additional payments afterwards. In this model a hard budget constraint

is denoted by @Ti2@Bi1

= 0. The inequality @Ti2@Bi1

> 0 on the other hand induces regional

opportunistic behaviour.

This positive model entails a soft budget constraint solely due to the order of ac-

tion. It doesn�t contribute to the explanation of the phenomenon why a government

cannot commit to a no-bailout-policy, but it helps to identify institutional arrange-

ments fostering the extent of the problem. The national government�s incentive to

grant payments follows from the maximisation of the welfare function

maxT2

nXi=1

[ui (Gi1) + �ui (Gi2) + wi (ci1) + �wi (ci2)] (16)

s.t.

ci1 = wi � Si (�) (17)

ci2 = (1 + �)Si (18)

Gi1 = Ti1 +Bi1 (19)

Gi2 = Ti2 + � iSi � (1 + ri)Bi1 (20)

nXi=1

Ti2 = � c

nXi=1

Si (21)

8i = 1; :::; n:

Using transfers T2 as instruments the capital tax rate � c is deduced through the

budget constraint (21). First order conditions of an interior solution are:

@ui@Gi2

=@uj@Gj2

8j 6= i: (22)

The national government chooses transfers to equate marginal utilities from the

provision of second period public goods across regions. This corresponds to the

common result in �scal federalism. This solution serves to uncover the national

government�s reaction to an increase in regional borrowing. The reaction function

7

is given by10dTi2dBi1

= 1 + ri: (23)

The national government increases transfers to borrowing jurisdictions. Ad-

ditional debt and interest payments due are balanced through additional vertical

grants. This guarantees equating marginal utility of public consumption. But the

national government must increase their tax rate on capital to �nance the bailout

payment. Using (9) and (11) net return � decreases, the marginal productivity of

capital ri increases and capital demand declines. Finally interest on debt increases.

This in turn can induce regions to lower their provision of public goods in the �rst

period too. To make up for these externalities, the national government may be

inclined to reduce grants to all regions j 6= i: The increase of interest payments ondebt a¤ects all regions including region i; hence also punishing the culprit region.

2.3 The regional government�s opportunistic behaviour with

tax competition

The soft budget constraint Turning to the �rst stage, regional governments

choose public good bundles. The regional government maximises the representative�s

household utility

maxBi1;� i

ui (Gi1) + �ui (Gi2) + wi (ci1) + �wi (ci2) (24)

s.t. eq. (17) ; (18) ; (19), (20), (21), (23). The �rst order conditions yield:

@ui@Gi1

� @ui@Gi2

=

�(1 + ri)�

dT �i2dBi2

�+drid� c

Bi1Pnj=1 Si

Xj

dT �j2dBi1

+� drid� iBi1 + S

iPnj=1 Si

d�d�cd�d� i

Xj

dT �j2dBi1

: (25)

The left hand side of eq. (25) denotes the marginal rate of substitution between

public good consumption in the �rst and second period. The right hand side denotes

the marginal rate of transformation and therefore the price of borrowing. The higher

the price, the lower public borrowing and public good provision in period one. As

discussed before, the national government increases transfers to regions with higher

debt. At the same time, transfers to all other regions are lowered to make up

10Cf. Breuillé et al. (2006), p. 234.

8

for negative externalities of capital taxation. The sign of the sum of all adjusted

transfers to the regionsP

j

dT �j2dBi1

is ambiguous. In accordance with Goodspeed (2002,

pp. 414f.), we consider the case of an increased sum of vertical grantsP

j

dT �j2dBi1

> 0.11

The deviation from the hard-budget-policy A hard-budget-policy requires

that no region can expect bailouts. In turn this a¤ords a national government not

reacting to higher de�cits by local governments: dTi2dBi1

=dTj2dBi1

= 0 8j 6= i: The price

of borrowing to regional governments is given by@ui@Gi1

�@ui@Gi2

= 1 + ri: Therefore three

e¤ects determine the softness of the budget constraint in eq. (25):

� The terms in brackets include the original price of borrowing 1 + ri and thereaction function of the national government. Following eq. (23) this e¤ect

vanishes. This e¤ect sincerely softens the budget constraint.

� The second term denotes the �common pool e¤ect�of taxation. Higher trans-fer payments induced by higher borrowing a¤ord an adjustment of the national

tax rate on capital. This in�uences capital demand, the net return on capital

diminishes while the before-tax interest rate ri increases. This vertically in-

duced e¤ect increases the price of borrowing and therefore hardens the budget

constraint.

� The third term denotes the �tax base sharing e¤ect�. The common tax base

induces regions to anticipate national government�s policy when setting taxes.

An increase of the national tax rate on capital reduces the tax base to all

regions. To make up for these externalities regions adjust tax rate in return.

Since regions compete for mobile capital an increase in tax revenue is either

achievable by higher tax rates or lower tax rates when broadening tax base.

Therefore tax setting depends on the �erceness of horizontal tax competition.

Tax setting in�uences tax base as well as the price of borrowing by changing

the before-tax interest rate ri: The sign of drid� iis undetermined. Therefore the

tax base sharing e¤ect on the price of borrowing is ambiguous.

All in all, the deviation from the hard budget depends on the extent of horizontal

tax competition.

11This assumption holds for all subsequently analysed cases.

9

2.4 The regional government�s opportunistic behaviour with-

out tax competition

Breuillé et al. (2006) uncover how tax competition e¤ects the softness of the regional

budget constraint. If capital is immobile, regions do not compete in taxes. Regional

or national tax on capital does not alter the demand for capital, the marginal pro-

ductivity of capital is non-varying drid� i

= drid�c

= 0: The net return on capital does

not react di¤erent to tax setting of regional or national authorities. Savings can-

not evade to other regions, that isd�d�cd�d�i

= 1: Incorporating the no tax competition

assumptions the price of borrowing is given by

@ui@Gi1

� @ui@Gi2

=

�(1 + ri)�

dT �i2dBi2

�+

SiPnj=1 Si

Xj

dT �j2dBi1

: (26)

Again the term in brackets vanishes due to eq. (23). Hence, Breuillé et al. (2006)

conclude that tax competition hardens the budget constraint if the right hand side

of eq. (25) is smaller than the right hand side of eq. (26):

drid� c

Bi1Pnj=1 Si

Xj

dT �j2dBi1

+� drid� iBi1 + S

iPnj=1 Si

d�d�cd�d� i

Xj

dT �j2dBi1

>SiPnj=1 Si

Xj

dT �j2dBi1

(27)

This simpli�es to Si > Bi1:

When regional debt is lower (higher) than the regional tax base, tax competition

hardens (softens) the budget constraint.12 As long as debt is fairly low, regions

cultivate their tax base. A burden on their tax base �induced through the �common

pool e¤ect�following higher debt �is to be avoided. But if debt exceeds tax base

excessive borrowing is bene�cial. Cultivating tax base is no priority. The result

suggests to settle for a debt ceiling in order to foster regional �scal discipline.

We now turn to the question, how a horizontal �scal equalisation scheme e¤ects

regional borrowing incentives within this framework. The paper takes account of

three di¤erent �scal equalisation schemes: revenue equalisation representing Ger-

many�s constitutional design; tax base equalisation representing Canada�s federal-

provincial �scal arrangements; and expenditure equalisation. We analyse these cases

in order of appearance.

12Cf. Breuillé et al. (2006), p. 235.

10

3 Tax revenue equalisation

Tax revenue equalisation is the key characteristic of the German �scal constitution.

We take account of the German formula in a simplifying manner within the model

discussed above.13 We consider a (partial) equalisation of capital tax revenue. In

the period of taxation all regions are involved in a horizontal scheme �2;fag also

applied by Köthenbürger (2002) and Baretti et al. (2002).14 Transfer payments for

region i in the second period amout to15

�i2;fag = �

�1

n

Pnj=1 � jS

j � � iSi�: (28)

The intensity of equalisation is given by rate � 2 [0; 1]. Sj denotes tax base of

region j and � j denotes the regional tax rate on capital. Horizontal grants for a

region eligible for compensation amounts to the intensity rate multiplied with the

deviation of the mean tax revenue 1n

Pnj=1 � jS

j from the region�s own tax revenue

� iSi: This net scheme is revenue neutral, that is

Pnj=1 �

j2;fag = 0 holds. Regions

with below average tax revenue receive grants, while regions with above average

tax revenue are due to contribute payments. The second period public budget of a

region is now given by

Gi2 = Ti2 + � iSi � (1 + ri)Bi1 + �

�1

n

Pnj=1 � jS

j � � iSi�: (29)

The sequence of action is unchanged. On the second stage, the national go-

vernment�s incentives to bail out borrowing regions is determined. The �rst order

conditions of an interior solution are given by

@ui@Gi2

=@uj@Gj2

8j 6= i:

The national government distributes vertical transfers to equate marginal utility

arising from public good consumption in the second period across regions. The reac-

tion function of the national government following an increase in public borrowing

of region k is obtained through implicit di¤erentiation of the optimality conditions.

13For a classi�cation of �scal horizontal equalisation schemes and their basic disincentive e¤ectssee Musgrave (1961).

14The subscript fag denotes �Finanzausgleich�, the German terminology of the horizontal �scalequalisation scheme.

15In the �rst period �i1;fag = 0 8i holds.

11

In accordance with eq. (23) the reaction function is unaltered.

dTi2dBi2

= 1 + ri:

The national government increases vertical grants to borrowing regions to ac-

count for their optimality conditions. Additional debt and interest payments due

are completely compensated through vertical grants.

3.1 Tax competition

The regional government maximises the representative household�s utility function

maxBi1;� i

ui (Gi1) + �ui (Gi2) + wi (ci1) + �wi (ci2)

s.t. (17) ; (18) ; (19), (29), (21), (23) : The �rst order conditions yield:

@ui@Gi1

� @ui@Gi2

=

�(1 + ri)�

dT �i2dBi2

�+drid� c

Bi1Pni=1 Si

Xj

dT �j2dBi1

+Si � dri

d� iBi1Pn

i=1 Si

d�d�cd�d� i

Xj

dT �j2dBi1

+�1� nn

d�d�cd�d� i

SiPni=1 Si

Xj

dT �j2dBi1

: (30)

We are thus able to state the following proposition:

Proposition 1 Tax revenue equalisation softens the budget constraint under taxcompetition. The price of borrowing decreases with an increase in the intensity of

equalisation rate �:

Proof. Subtracting the price of borrowing faced by region i if involved in a tax

revenue equalisation scheme from the price of borrowing without �scal equalisation

by eq. (25) leaves only a single term. Due to 1 � n < 0 the term is negative and

therefore lowers the price of borrowing. The lower price induces regions to increase

their provision of public goods in the �rst period through an increase in de�cit. The

intensity rate � ampli�es the e¤ect, independent of the region being a net payer or

12

net recipient. Di¤erentiating the right hand side of eq. (30) with respect to � yields:

1� nn

d�d�cd�d� i

SiPni=1 Si

Xj

dT �j2dBi1

< 0

8 n > 1:

The price of borrowing declines with an increase in the intensity rate.

This result shows that a tax revenue scheme has negative e¤ects on �scal dis-

cipline. Tax �nancing public goods is complicated because additional revenue is

�taxed�by the equalisation scheme. To �nance one additional Euro of public goods,

tax revenue must increase by 11�� > 1 Euro (neglecting the e¤ect on average tax

revenue). Hence, we can conclude: the price of public goods provided through

additional taxation increases if states are involved in a tax revenue equalisation

scheme. Regions substitute period two public good provision, �nanced by taxation

(and grants), with increased public provision in the �rst period, which is �nanced by

additional borrowing. The �tax�on tax revenue is responsible for this disincentive

e¤ect. Increased vertical grants in the second period makes repaying debt burdens

even more a¤ordable.

Public goods are preferably �nanced by bailout payments from the national go-

vernment. An increase of the regional tax rate within an tax revenue equalisation

scheme results in higher payments for net payer or lower (horizontal) grants for

net recipients. The region acts opportunistic when setting low tax rates and bor-

rowing more. Because the national government cannot deny payments ex post to

equate marginal utility from public consumption, regions are induced to miss out

on �scal discipline. The price of borrowing decreases with stronger horizontal tax

competition, that is the higher the capital elasticity.

3.2 Without tax competition

To account for German circumstances, the analysis neglecting (strong) tax compe-

tition is of greater importance. If capital is immobile, regions do not compete in

taxes. Regional or national tax on capital does not alter the demand for capital,

the marginal productivity of capital is non-varying and drid� i= dri

d�c= 0 holds. The

net return on capital does not react di¤erent to tax setting of regional or national

authorities, that isd�d�cd�d�i

= 1 holds. Incorporation the no tax competition assumptions

13

into eq. (30) reduces the price of borrowing to

@ui@Gi1

� @ui@Gi2

=

�(1 + ri)�

dT �i2dBi2

�+1

n

Si

1n

Pni=1 Si

Xj

dT �j2dBi1

+1

n

��1� nn

�Si

1n

Pni=1 Si

Xj

dT �j2dBi1

: (31)

We are thus able to state the following proposition:

Proposition 2 Tax revenue equalisation softens the budget constraint without taxcompetition. The price of borrowing decreases with an increase in the intensity of

equalisation rate �, and with decrease in the (relative) tax base.

Proof. Subtracting the price of borrowing faced by region i if involved in a

tax revenue equalisation scheme according to eq. (31) from the price of borrowing

without �scal equalisation by eq. (26) leaves only a single term. Due to 1 � n < 0the term is negative and therefore lowers the price of borrowing. The lower price

induces regions to increase their provision of public goods in the �rst period through

de�cit spending.�1 +

� (1� n)n

�SiPni=1 Si

Xj

dT �j2dBi1

<SiPnj=1 Si

Xj

dT �j2dBi1

� (1� n) < 0:

The intensity rate � ampli�es the e¤ect, independent of the region being a net payer

or net recipient. Di¤erentiating the right hand side of eq. (31) with respect to �

yields:1� nn

SiPni=1 Si

Xj

dT �j2dBi1

< 0:

Both conditions hold for 8 n > 1.A marginal increase in the relative tax base increases the price of borrowing.

Di¤erentiating the right hand side of eq. (31) with respect to Si1n

Pni=1 Si

yields

1

n

�1 + �

1� nn

�Xj

dT �j2dBi1

> 0;

which is positive, since 1 + �1�nn= (1��)n+�

n> 0 holds for 0 < � < 1.

Tax revenue equalisation exhibits negative e¤ects on �scal discipline independent

of tax competition. The regions substitute the expensive provision of public goods

14

in the second period �nanced by taxation by increased borrowing to �nance more

public goods in the �rst period. Borrowing incentives rise with the intensity rate of

the equalisation scheme.

A relative large tax base alleviates borrowing incentives. Although any region

can externalise the debt burden on the national tax payer, a rich region with a large

tax base will bear a larger share of the national tax burden than a poor region. This

can analytically demonstrate the lower borrowing incentives of stronger economic

powers.

Summing up, the analysis yields a debt incentive function

Bi1 = Bi1

�i;

Si

1n

Pnj=1 Sj

!: (32)

Theoretical reasoning suggests the partial derivative with respect to the intensity

rate is positive and the partial derivative with respect to the relative tax base is neg-

ative. This model can o¤er theoretical support for Rodden�s (2003) statement:�[:::]

Germany�s complex, interdependent, collaborative style of federalism tends to di-

lute �scal accountability and soften budget constraints.�(p. 164). And further on:

�Perhaps the most important source of soft Land-level budget constraints in Ger-

many is the Constitution itself. By simultaneously creating wide-ranging Land-level

administrative autonomy, guaranteeing the equivalence of living conditions, and en-

trenching strong representation of the Länder in the federal policymaking process,

the Constitution creates hurdles for the proper functioning of several �scal discipline

mechanisms.�(p. 181).

4 Tax base equalisation

Tax base equalisation is the key characteristic of the Canadian and to some extent

of the Swiss federal �scal arrangements. We take account of the Canadian formula

in a simplifying manner within the model discussed above. We consider a (partial)

equalisation of capital tax base. In the period of taxation all regions are involved

in a horizontal scheme �2;wka also applied by Köthenbürger (2002) and Bucovetsky

15

and Smart (2006).16 Transfer payments for region i in the second period amout to17

�i2;wka = ���

�1

n

Pnj=1 S

j � Si�: (33)

�� represents a mean (regional) taxation rate:

�� =

Pnj=1 � jS

jPnj=1 S

j: (34)

The institutional scheme equates di¤erences in tax revenue resulting from di¤er-

ent tax bases.18 The tax base serves an indicator of economic power. Regions with

below (above) average tax bases are eligible (liable) to horizontal payments These

horizontal payments amount to the product comprising the mean taxation rate, the

intensity of equalisation rate, and the deviation of the region�s tax base from the

mean tax base. A region having no own tax base is equipped with an average tax

revenue. The equalisation scheme is revenue neutral, that isPn

j=1 �j2;wka = 0 holds.

The underlying economic-philosophic implication of this systems has been de-

scribed by Musgrave (1961, p. 103): �[...] the societies of each state should be

permitted to determine their own levels of �scal activity, but [...] the central govern-

ment should equalize the �scal opportunities of the various states, or the potential

levels which they might achieve with their own action.�

The regional public budget in the second period is now given by

Gi2 = Ti2 + � iSi � (1 + ri)Bi1 + ���

�1

n

Pnj=1 S

j � Si�: (35)

The incentive of the national government to bailout borrowing regions is un-

altered. Eq. (23) maintains its validity. To uncover borrowing incentives of the

regional governments we turn immediately to the �rst stage, in which they decide

upon the provision of public goods.

16The subscript wka denotes the abbreviation of �Wirtschaftskraftausgleich�, the German ter-minology of equalising economic power.

17In the �rst period �i1;wka = 0 8i holds.18Cf. Smart (2005) for a short description of the Canadian federal-provincial �scal arrangements

act.

16

4.1 Tax competition

The regional government maximises the representative household�s utility

maxBi1;� i

ui (Gi1) + �ui (Gi2) + wi (ci1) + �wi (ci2)

s.t. (17) ; (18) ; (19), (35), (21), (23). The �rst order conditions yield:

@ui@Gi1

� @ui@Gi2

= (1 + ri)�dT �i2dBi2

+drid� c

Bi1Pni=1 Si

Xj

dT �j2dBi1

+Si � dri

d� iBi1Pn

i=1 Si

d�d�cd�d� i

Xj

dT �j2dBi1

+�

264Si 1nPnj=1 S

j � Si�Pnj=1 S

j�2 d�

d�cd�d� i

Xj

dT �j2dBi1

375 : (36)

Eq. (36) shows the price of borrowing faced by region i. The higher the right

hand side, the higher the cost of providing public goods in the �rst period. If the

right hand side is low, the incentive to borrow is high. Therfore, the right hand side

indicates the softness of the budget constraint. Subsequently we analyse the softness

in comparison to the base case without �scal equalisation and in comparison to the

tax revenue equalisation scheme.

Base case without �scal equalisation In comparison to the base case (eq.

(25)) eq. (36) exhibits an additional term which can be of positive or negative sign.

Therefore borrowing incentives may decrease or increase. The sign depends on the

economic power of a region, which also decides upon the status of a region being net

payer or net recipient. If a region exhibits a tax base Si below average 1n

Pnj=1 S

j

(net recipient), the sign becomes positive. This region is exposed to a higher price

of borrowing. If a region exhibits a tax base above average (net payer), the sign

becomes negative. This region is exposed to a lower price of borrowing.19

Proposition 3 Tax base equalisation hardens (softens) the budget constraint fornet recipients (payers) under tax competition. The intensity rate � ampli�es the

respective e¤ect.

19Interestingly, the European Stability and Growth Pact (Code of Conduct (2005)) envisagesa mid-term budget rule for member states, which corresponds to the incentive scheme described.While low indebted countries and countries with a high growth potential are conceded with a de�cittarget of up to one percent of GNP, high indebted countries and countries with a lower growthpotential are supposed to have a budget �close-to-balance or in surplus�.

17

Proof. Tax base equalisation softens the budget constraint if the price of borrowing

by eq. (36) is lower than the price of borrowing without �scal equalisation by eq.

(25) :

�

264Si 1nPnj=1 S

j � Si�Pnj=1 S

j�2 d�

d�cd�d� i

Xj

dT �j2dBi1

375 < 0

Si >1

n

Pnj=1 S

j

The intensity rate softens the budget constraint if the right hand side of eq. (36)

depends negatively on � and vice versa. The di¤erentiated right hand side with

respect to � is negative, only if

Si >1

n

Pnj=1 S

j

holds. Both conditions are met by net payers. The reverse holds for net recipients.

They are exposed to a lower incentive to borrow.

The reason, that a region with a low tax base is exposed to a harder budget

constraint, lies in the horizontal grants received from high tax base regions. As net

recipients they are exposed to higher borrowing costs,.which leads to a lower debt

and a lower provision of public goods in the �rst period. Debt �nances additional

public goods in the �rst, taxation �nances public goods in the second period. But

�rst period debt e¤ects on the one hand side second period horizontal transfer

payments, due to the �nancing of bailout payments via the national tax on capital.

Higher bailouts resulting in a higher national tax rate reduces the net return on

capital a¤ecting all tax bases in the federation. This happens independently of the

status as net payer or net recipient.

On the other hand side taxation in the second period e¤ects transfer payments

di¤erently depending on the status. Net recipients can increase the mean tax rate,

which they appreciate due to higher grants. Net payers su¤er from an increase in the

mean tax rate due to higher payments. Therefore, net recipients can easily �nance

public goods in the second period because higher taxation is not �taxed away�as

under the tax revenue scheme. On the contrary, taxation is rewarded by additional

grants from the rich regions. Net recipients are induced to higher tax setting and

lower borrowing.

Net payers can lower the mean tax rate through a reduction in their own tax

rate. They appreciate a low mean tax rate due to lower payments. Therefore net

18

payers are induced to lower tax setting and higher borrowing. The e¤ect works

solely through the mean tax rate. Should the regions falsely anticipate the mean

tax rate being constant, no deviating borrowing incentives in comparison to the base

case without �scal equalisation arise.

The tax base equalisation scheme strengthens �scal discipline in �poor�economic

regions, due to their reliance upon horizontal grants from richer regions. This re-

minds of the work of Huber and Runkel (2008). They demonstrate the advantage

of country-speci�c budget rules in a model of asymmetric information. Accordingly,

net recipients should be imposed to harder budget rules. The tax base equalisation

scheme implicitly induces these incentives.

Base case tax revenue equalisation Net recipients are exposed to lower bor-

rowing incentives in a tax base equalisation scheme in comparison to the base case

without a �scal equalisation scheme. Because the tax revenue equalisation exhibits

higher borrowing incentives independently of the status of a region, net recipients

are also exposed to lower borrowing incentives in a tax base equalisation scheme in

comparison to the tax revenue scenario. Which equalisation scheme yields higher

borrowing incentives for net payers is yet to be named.

We are able to state the following proposition:

Proposition 4 A tax base equalisation scheme hardens the budget constraint in

comparison to a tax revenue equalisation scheme with tax competition independently

of the status of a region as net payer or net recipient.

Proof. Tax base equalisation hardens the budget constraint, if the price of bor-

rowing according to eq. (36) is higher than the price of borrowing according to eq.

(30):

�Si1n

Pnj=1 S

j � Si�Pnj=1 S

j�2 d�

d�cd�d� i

Xj

dT �j2dBi1

> �1� nn

d�d�cd�d� i

SiPni=1 Si

Xj

dT �j2dBi1

SiPnj=1 S

j< 1:

As long as a region cannot attract all mobile capital, borrowing incentives are

always lower in a tax base equalisation scheme.

The tax base equalisation scheme is the incentive friendly alternative to the

tax revenue equalisation scheme. The reason is found in the missing �tax�on tax

19

revenue. Additional tax revenue remains ceteris paribus in the source region. Still,

the e¤ect on the mean tax rate does distort borrowing incentives.

4.2 Without tax competition

To account for German circumstances, the analysis neglecting tax competition is of

greater interest. But due to the results obtained above, which were not dependent

on the extent of tax competition, all insights remain unchanged. Without tax com-

petition drid� i= dri

d�c= 0 and d�

d�c= d�

d� iholds. The price of borrowing according to eq.

(36) simpli�es to:

@ui@Gi1

� @ui@Gi2

= (1 + ri)�dT �i2dBi2

+SiPni=1 Si

Xj

dT �j2dBi1

+�

264Si 1nPnj=1 S

j � Si�Pnj=1 S

j�2 X

j

dT �j2dBi1

375 : (37)

Base case without �scal equalisation

Proposition 5 Tax base equalisation hardens (softens) the budget constraint fornet recipients (payers) without tax competition. The intensity rate � ampli�es the

respective e¤ect.

Proof. Tax base equalisation softens the budget constraint if the price of borrowing

by eq. (37) is lower than the price of borrowing without �scal equalisation by eq.

(26):

�

264Si 1nPnj=1 S

j � Si�Pnj=1 S

j�2 X

j

dT �j2dBi1

375 < 0

Si >1

n

Pnj=1 S

j (38)

The intensity rate softens the budget constraint if the right hand side of eq. (37)

depends negatively on � and vice versa. The di¤erentiated right hand side with

respect to � is negative, only if

Si >1

n

Pnj=1 S

j

20

holds. Both conditions are met by net payers. The reverse holds for net recipients.

They are exposed to a lower incentive to borrow.

Base case tax revenue equalisation The comparison of the incentive structure

of a tax base equalisation to a tax revenue equalisation without tax competition

yields the following proposition:

Proposition 6 A tax base equalisation scheme hardens the budget constraint in

comparison to a tax revenue equalisation scheme without tax competition indepen-

dently of the status of a region as net payer or net recipient.

Proof. Tax base equalisation hardens the budget constraint, if the price of bor-

rowing according to eq. (37) is higher than the price of borrowing according to eq.

(31):

�Si1n

Pnj=1 S

j � Si�Pnj=1 S

j�2 X

j

dT �j2dBi1

> �1� nn

SiPni=1 Si

Xj

dT �j2dBi1

SiPnj=1 S

j< 1:

As long as a region cannot attract all mobile capital, borrowing incentives are

always lower in a tax base equalisation scheme.

These results allow for a ranking of borrowing incentives. Figure (1) illustrates

the softness of the budget constraint with respect to the intensity of equalisation

rate �: A �hard budget�exhibits low, and a �soft budget�exhibits high borrowing

incentives. The graphs distinguish between net payers and net recipients.

Tax revenue softens the budget constraint in comparison to the base case with-

out �scal equalisation. The e¤ect increases in the intensity of equalisation rate. Net

recipients are exposed to higher borrowing incentives, due to the circumstance that

bailouts are �nanced to a greater share by the �rich�regions. Tax base equalisation

exposes net recipients to lower borrowing incentives and net payers to higher bor-

rowing incentives in comparison to the base case without �scal equalisation. The

budget constraint under tax base equalisation is in any case harder than under tax

revenue equalisation.

21

Figure 1: Fiscal equalisation and the soft budget constraint

1

tax base eq. ­ net recipient

tax base eq. ­ net payer

tax revenue eq. ­ net payer

tax revenue eq. – net recipient

HardBudget

SoftBudget

0intensity rate

base case without eq.

5 Expenditure equalisation

Groneck and Plachta (2008) o¤er an alternative horizontal equalisation scheme

which intends to establish a debt brake in �scal equalisation. This part assesses

the theoretical reasoning of the scheme. The advisory council at the Department

of the Treasury in Germany discussed the possibility to reform budget rules and

prohibit the development of budgetary crisis by levying a due from all regional go-

vernments proportional to their de�cit.20 This corresponds to a tax on debt.21 A

further proposal by Söllner (2000) suggests to accompany the Maastricht criteria on

state level by a �scal equalisation scheme incorporating not only tax revenue but

also borrowing potential (p. 17).

On the basis of these inspirations we will subsequently analyse, how the inclusion

of debt in a German type of equalisation scheme e¤ects borrowing incentives. Equal-

ising all state originating �nancial resources corresponds to equalising expenditure,

hence we shall call this scheme expenditure equalisation for the ease of writing.22

The mechanism of the rule is the following: states with above average de�cit �at

equal tax revenue �are liable to higher payments (if net payer) or have to abstain

from a share of grants (if net recipient). This analysis is focused on the question, in

how far the soft-budget-phenomenon is a¤ected.

20Cf. Wissenschaftlicher Beirat beim Bundesministerium der Finanzen (2005), p. 25.21Stähler (2007) discusses a similar proposal, whereby increased borrowing is necessarily accom-

panied by increased taxation.22This is technically not correct, due to the neglection of vertical grants.

22

A transfer system �sfa is integrated into the base model presented above.23 In

contrast do the schemes already analysed, it is now necessary to consider �rst period

equalisation. First period transfers �at equal exogenous revenue Ti1 in all regions

i = 1; :::; n �are given by

�i1;sfa = �

�1

n

Pnj=1D1j �D1i

�= �

�1

n

Pnj=1B1j �B1i

�; (39)

whereas de�cit D1 corresponds to the debt level at the end of the �rst period B1.

Thus, in the �rst period de�cits are (partially) equalised across states. Above aver-

age de�cits are �taxed�with rate �.

Transfer payments in the second period are given by

�i2;sfa = �

�1

n

Pnj=1

�� jS

j +D2j

���� iS

i +D2i

��= �

�1

n

Pnj=1

�� jS

j �B1j���� iS

i �B1i��; (40)

whereas budget must be balanced after the second period with no debt being passed

on. Second period de�cit is therefore equal to the negative debt level at the end of

period one D2k = �B1k: 8k: In the second period expenditure is partially equalised.The public budget constraint in period one amounts to:

Gi1 = Ti1 +Bi1 + �

�1

n

Pnj=1B1j �B1i

�: (41)

The public budget constraint in period two amounts to:

Gi2 = Ti2 (Bi1) + � iSi � (1 + ri)Bi1

+ �

�1

n

Pnj=1

�� jS

j �B1j���� iS

i �B1i��: (42)

The incentive of the national government to bail out borrowing regions is un-

altered. Eq. (23) maintains its validity. To uncover borrowing incentives of the

regional governments we turn immediately to the �rst stage, in which they decide

upon the provision of public goods.

23The subscript sfa denotes the abbreviation of �Schuldenbremse im Finanzausgleich�, a Ger-man name advertising the scheme.

23

5.1 Tax competition

The regional government maximises the representative household�s utility function

maxBi1;� i

ui (Gi1) + �ui (Gi2) + wi (ci1) + �wi (ci2)

s.t. (17) ; (18) ; (41), (42), (21), (23). The �rst order conditions yield:

@ui@Gi1

� @ui@Gi2

=n

(1� �)n+ �

"(1 + ri)�

dT �i2dBi2

+drid� c

Bi1Pni=1 Si

Xj

dT �j2dBi1

+Si � dri

d� iBi1Pn

i=1 Si

d�d�cd�d� i

Xj

dT �j2dBi1

+�1� nn

d�d�cd�d� i

SiPni=1 Si

Xj

dT �j2dBi1

+ �1� nn

#: (43)

In comparison to the base case without �scal equalisation two terms inside the

square brackets and one factor outside have emerged. The second to last term in-

side the square brackets replicates the negative tax revenue equalisation e¤ect, as

described in section 3. Tax revenue equalisation increases the price of tax �nanced

provision of public goods. Regions substitute expensive second period public con-

sumption by �rst period consumption �nanced by borrowing. The last term in

square brackets is also negative, due to n > 1. This e¤ect results from repaying

debt in the second period. High de�cits in the �rst period result in low possible

expenditure in the second period. This e¤ect is in turn partially o¤set due to higher

demands to the equalisation system resulting in higher grants for net recipient or

lower payments for net payer states. The square brackets are multiplied with the

factor n(1��)n+� > 1 8 n > 1; 8 0 < � < 1. This last e¤ect increases the price of

borrowing.

The e¤ect of the intensity of equalisation rate � is ambiguous.

Proposition 7 Expenditure equalisation hardens (softens) the budget constraint iftax competition e¤ects are strong (weak).

Proof. Expenditure equalisation hardens the budget constraint, if the price of

borrowing by eq. (43) is higher than the price of borrowing without equalisation by

24

eq. (25). This is true for"drid� c

� drid� i

d�d�cd�d� i

#Bi1Pni=1 Si

Xj

dT �j2dBi1

> 1: (44)

Two e¤ects determine the result: the �common pool e¤ect�through drid�cand the

�tax base sharing e¤ect�through drid� i

d�d�cd�d�i

: The former re�ects the in�uence of higher

vertical grants (bailouts) due to increased borrowing, which a¤ects the national tax

rate on capital. This in turn a¤ects capital demand, net return on capital decreases,

the before-tax interest rate ri increases. This vertical e¤ect increases the price of

borrowing, thus hardens the budget constraint. The latter e¤ect is ambiguous due

to multiple interactions, that is no clear cut prediction is possible, how regions alter

their tax rates in expectation of the adjustment of the national tax rate on capital.24

If horizontal tax externalities are strong, that is an increase in the regional tax rate

results in a high out�ow of tax base and a large increase in the before-tax interest

rate ri, the region will want to choose a moderate increase or even a reduction

of their tax rate. Then drid�c� dri

d� i

d�d�cd�d�i

> 0 holds and the debt brake would work. If

horizontal tax competition is weak, that is an increase in the regional tax rate results

in a fairly low out�ow of tax base and a moderate increase in the before-tax interest

rate ri, the region could be induced to choose a strong increase of their tax rate.

Then drid� i

d�d�cd�d�i

> drid�c

holds and the debt brake falls short of its name. As long as the

�tax base sharing e¤ect�doesn�t compensate the �common pool e¤ect�expenditure

equalisation hardens the budget constraint.

5.2 Without tax competition

Without tax competition holds. drid� i= dri

d�c= 0 and d�

d�c= d�

d� i. The price of borrowing

according to eq. (43) simpli�es to:

@ui@Gi1

� @ui@Gi2

=n

(1� �)n+ �

�(1 + ri)�

dT �i2dBi2

+

�1 + �

1� nn

�SiPni=1 Si

Xj

dT �j2dBi1

+ �1� nn

�: (45)

24See section 2.3.

25

Proposition 8 Expenditure equalisation softens the budget constraint without taxcompetition.

Proof. Eq. (44) simpli�es to

[0] � Bi1Pni=1 Si

Xj

dT �j2dBi1

> 1:

This is a contradiction. Therefore, expenditure equalisation cannot harden the bud-

get constraint.

6 Conclusion

This paper tackles the question which e¤ects a �scal equalisation scheme exerts on

debt policy of sovereign regional governments. Tax revenue equalisation softens the

budget constraint of a government independent of the status as net payer or net

recipient. Borrowing incentives increase with the intensity rate and decrease with

the relative tax base. Tax base equalisation hardens the budget constraint for net

recipients in comparison to no �scal equalisation. In turn it softens the budget

constraint for net payers. Incentives to debt �nance public goods are always lower

under tax base equalisation in comparison to tax revenue equalisation. All results

are independent of the intensity of tax competition. A third equalisation scheme,

comprising tax revenue and de�cits, can serve as a scheme to reduce borrowing

incentives if horizontal tax competition is strong. Empirical testing is left for further

research.

References

Baretti, C., Huber, B. and Lichtblau, K. (2002). A tax on tax revenue: The incentive

e¤ects of equalizing transfers: Evidence from Germany, International Tax and

Public Finance 9(6): 631�649.

Breuillé, M.-L., Madiès, T. and Taugourdeau, E. (2006). Does tax competition

soften regional budget constraint?, Economic Letters 90(2): 230�236.

Breuillé, M.-L., Madiès, T. and Taugourdeau, E. (2007). Funding source and soft

budget constraint, EconomiX Working Paper, 2007-12.

26

Bucovetsky, S. and Smart, M. (2002). The e¢ ciency consequences of local revenue

equalization: Tax competition and tax distortions, CESifo Working Paper Se-

ries 767.

Bucovetsky, S. and Smart, M. (2006). The e¢ ciency consequences of local revenue

equalization: Tax competition and tax distortions, Journal of Public Economic

Theory 8(1): 119�144.

Code of Conduct (2005). Speci�cations on the implementation of the stability and

growth pact and guidelines on the format and content of stability and conver-

gence programmes, Europäische Kommission - ECOFIN Council.

Goodspeed, T. J. (2002). Bailouts in a federation, International Tax and Public

Finance 9(4): 409�421.

Groneck, M. and Plachta, R. C. (2008). Die Schuldenbremse im Finanzausgleich -

Eine alternative Budgetregel für die Bundesländer, Zeitschrift für Wirtschaft-

spolitik 3: 259�284.

Huber, B. and Runkel, M. (2008). Interregional distribution and budget institutions

under asymmetric information, Journal of Public Economics 92: 2350�2361.

Keen, M. J. and Kotsogiannis, C. (2002). Does federalism lead to excessively high

taxes?, American Economic Review 92(1): 363�370.

Kornai, J. (1979). Resource-constrained versus demand-constrained systems, Econo-

metrica 47(4): 801�819.

Köthenbürger, M. (2002). Tax competition and �scal equalization, International

Tax and Public Finance 9(4): 391�408.

Musgrave, R. A. (1961). Approaches to a �scal theory of political federalism, Pub-

lic Finances: Needs, Sources, and Utilization, National Bureau of Economic

Research, Princeton University Press, Princeton.

Oates, W. (2005). Towards a second generation theory of �scal federalism, Interna-

tional Tax and Public Finance 12(4): 349�373.

Qian, Y. and Roland, G. (1998). Federalism and the soft budget constraint, Amer-

ican Economic Review 88(5): 1143�1162.

27

Rodden, J. (2003). Reviving leviathan: Fiscal federalism and the growth of govern-

ment, International Organization 57(4): 695�729.

Rodden, J., Eskeland, G. S. and Litvack, J. (2006). Fiscal decentralization and the

challenge of hard budget constraints, National Tax Journal 59(2): 389�396.

Smart, M. (2005). Some notes on equalization reform, Working Paper, Department

of Economics, University of Toronto.

Söllner, F. (2000). Die Umsetzung der Maastricht-Kriterien zur Staatsverschuldung

im föderalen System der Bundesrepublik Deutschland, ifo-Schnelldienst 10: 8�18.

Stähler, N. (2007). Taxing de�cits to restrain government spending and foster cap-

ital accumulation, Discussion Paper, Series 1: Economic Studies, Deutsche

Bundesbank.

Vigneault, M. (2007). Grants and soft budget constraints, in R. Boadway and

A. Shah (eds), Intergovernmental Fiscal Transfers, Principles and Practice,

The World Bank, Washington, D.C., chapter 5, pp. 133�171.

Wilson, J. D. (1986). A theory of interregional tax competition, Journal of Urban

Economics 19(3): 296�315.

Wissenschaftlicher Beirat beim Bundesministerium der Finanzen (2005). Haushalt-

skrisen im Bundesstaat, Schriftenreihe des Bundesministeriums der Finanzen

78.

Zodrow, G. and Mieszkowski, P. (1986). Pigou, Tiebout, property taxation, and the

underprovision of local public goods, Journal of Urban Economics 19(3): 356�370.

28

FiFo-CPE Discussion Papers Finanzwissenschaftliche Diskussionsbeiträge

Eine Schriftenreihe des Finanzwissenschaftlichen Forschungsinstituts an der Universität zu Köln; ISSN 0945-490X.

Die Beiträge ab 1998 stehen auch als kostenloser Download zur Verfügung unter: http:// fifo-koeln.de oder http://cpe.uni-koeln.de.

Discussions Papers from 1998 onwards can be downloaded free of charge from: http:// fifo-koeln.de or http://cpe.uni-koeln.de.

95-1 Scholl, R.: Verhaltensanreize der Abwasserabgabe: eine Untersuchung der Tarifstruktur der Abwasserabgabe. 6,50 EUR.

95-2 Kitterer, W: Intergenerative Belastungsrechnungen („Generational Accounting“) - Ein Maßstab für die Belas-tung zukünftiger Generationen? 7,50 EUR.

96-1 Ewringmann, D./Linscheidt, B./Truger, A.: Nationale Energiebesteuerung: Ausgestaltung und Aufkommens-verwendung. 10,00 EUR.

96-2 Ewringmann, D./Scholl. R.: Zur fünften Novellierung der Abwasserabgabe; Meßlösung und sonst nichts? 7,50 EUR.

97-1 Braun, St./Kambeck, R.: Reform der Einkommensteuer. Neugestaltung des Steuertarifs. 7,50 EUR.

97-2 Linscheidt, B./Linnemann, L.: Wirkungen einer ökologi-schen Steuerreform – eine vergleichende Analyse der Modellsimulationen von DIW und RWI. 5,00 EUR.

97-3 Bizer, K./Joeris, D.: Bodenrichtwerte als Bemessungs-grundlage für eine reformierte Grundsteuer. 7,50 EUR.

98-1 Kitterer, W.: Langfristige Wirkungen öffentlicher Investi-tionen - theoretische und empirische Aspekte. 6,00 EUR.

98-2 Rhee, P.-W.: Fiskale Illusion und Glory Seeking am Beispiel Koreas (1960-1987). 5,00 EUR.

98-3 Bizer, K.: A land use tax: greening the property tax system. 5,00 EUR.

00-1 Thöne, M.: Ein Selbstbehalt im Länderfinanzausgleich?. 6,00 EUR.

00-2 Braun, S., Kitterer, W.: Umwelt-, Beschäftigungs- und Wohlfahrtswirkungen einer ökologischen Steuerreform: eine dynamische Simulationsanalyse unter besonderer Berücksichtigung der Anpassungsprozesse im Übergang. 7,50 EUR.

02-1 Kitterer, W.: Die Ausgestaltung der Mittelzuweisungen im Solidarpakt II. 5,00 EUR.

05-1 Peichl, A.: Die Evaluation von Steuerreformen durch Simulationsmodelle. 8,00 EUR.

05-2 Heilmann, S.: Abgaben- und Mengenlösungen im Klima-schutz: die Interaktion von europäischem Emissions-handel und deutscher Ökosteuer. 8,00 EUR.

05-3 Fuest, C., Peichl, A., Schaefer, T.: Dokumentation FiFo-SiM: Integriertes Steuer-Transfer-Mikrosimulations- und CGE-Modell. 8,00 EUR.

06-1 Fuest, C., Peichl, A., Schaefer, T.: Führt Steuervereinfa-chung zu einer „gerechteren“ Einkommensverteilung? Eine empirische Analyse für Deutschland. 6,00 EUR.

06-2 Bergs, C., Peichl, A.: Numerische Gleichgewichtsmodelle - Grundlagen und Anwendungsgebiete. 6,00 EUR.

06-3 Thöne, M.: Eine neue Grundsteuer – Nur Anhängsel der Gemeindesteuerreform? 6,00 EUR.

06-4 Mackscheidt, K.: Über die Leistungskurve und die Besol-dungsentwicklung im Laufe des Lebens. 6,00 EUR.

06-5 Fuest, C., Peichl, A., Schaefer, T.: Does tax simplification yield more equity and efficiency? An empi-rical analysis for Germany. 6,00 EUR.

06-6 Fuest, C., Peichl, A., Schaefer, T.: Die Flat Tax: Wer gewinnt? Wer verliert? Eine empirische Analyse für Deutschland. 6,00 EUR.

06-7 Kitterer, W., Finken, J.: Zur Nachhaltigkeit der Länder-haushalte – eine empirische Analyse. 6,00 EUR.

06-8 Bergs, C., Fuest, C., Peichl, A., Schaefer, T.: Reformop-tionen der Familienbesteuerung: Aufkommens-, Vertei-lungs- und Arbeitsangebotseffekte. 6,00 EUR.

06-9 Ochmann, R., Peichl, A.: Measuring distributional effects of fiscal reforms. 10,00 EUR.

06-10 Peichl, A., Schaefer, T.: Documentation FiFoSiM: Integrated tax benefit microsimulation and CGE model. 8,00 EUR.

06-11 Peichl, A., Schaefer, T., Scheicher, C.: Measuring Rich-ness and Poverty. A micro data application to Germany and the EU-15. 6,00 EUR.

07-1 Fuest, C., Mitschke, J., Peichl, A., Schaefer, T.: Wider die Arbeitslosigkeit der beruflich Geringqualifizierten: Entwurf eines Kombilohn-Verfahrens für den Niedrig-lohnsektor. 8,00 EUR.

07-2 Groneck, M. Plachta, R.: Eine natürliche Schuldenbrem-se im Finanzausgleich. 6,00 EUR.

07-3 Kitterer, W.: Bundesstaatsreform und Zukunft der Finanzverfassung. 8,00 EUR.

07-4 Brenneisen, F., Peichl, A.: Dokumentation des Wohl-fahrtsmoduls von FiFoSiM. 6 EUR.

07-5 Brenneisen, F., Peichl, A.: Empirische Wohlfahrtsmes-sung von Steuerreformen. 6 EUR.

07-6 Fuest, C., Peichl, A., Schaefer, T.: Is a Flat Tax politically feasible in a grown-up Welfare State? 6,00 EUR.

07-7 Groneck, M., Plachta, R.: Simulation der Schuldenbrem-se und der Schuldenschranke für die deutschen Bundes-länder. 12,00 EUR.

07-8 Becker, J., Fuest, C.: Tax Enforcement and Tax Havens under Formula Apportionment. 6,00 EUR

07-9 Fuest, C., Peichl, A.: Grundeinkommen vs. Kombilohn: Beschäftigungs- und Finanzierungswirkungen und Un-terschiede im Empfängerkreis. 6,00 EUR.

08-1 Thöne, M.: Laffer in Luxemburg: Tankverkehr und Steueraufkommen im Großherzogtum. 6,00 EUR.

08-2 Fuest, C., Thöne, M.: Staatsverschuldung in Deutsch-land: Wende oder Anstieg ohne Ende? 6,00 EUR.

08-3 Becker, J., Peichl, A., Rincke, J.: Politicians’ outside earnings and electoral competition. 6,00 EUR.

08-4 Paulus, A., Peichl, A.: Effects of flat tax reforms in Western Europe on equity and efficiency. 6,00 EUR.

08-5 Peichl, A., Schaefer, T.: Wie progressiv ist Deutschland? Das Steuer- und Transfersystem im europäischen Ver-gleich. 6,00 EUR.

08-6 Peichl, A., The benefits of linking CGE and Microsimulation Models - Evidence from a Flat Tax analysis. 6,00 EUR.

08-7 Groneck, M.: A Golden Rule of Public Finance or a Fixed Deficit Regime? Growth and Welfare Effects of Budget Rules. 6,00 EUR.